A method for generating a flight buffer zone of an unmanned aerial vehicle based on homogeneous rules

Abstract

The present invention relates to the fields of UAV technology and power system inspection, and particularly to a method for generating a UAV flight buffer based on homogeneous rules. The present invention provides a method for generating a UAV flight buffer based on homogeneous rules. The key points are that the UAV system acquires tower information and coordinates, and the point cloud system acquires the environment around the tower; a flight buffer generation model is constructed according to the initialization parameters and homogeneous rules, and the model aims to generate a complete closed buffer; first, the acquired geographical coordinates are converted into projected coordinates, the buffer point coordinates are generated using the model, and finally the projected coordinates are converted back into geographical coordinates. The purpose is to use this method to generate a UAV buffer using homogeneous rules, considering the geometric distances and attribute characteristics of all instances within the target area, so as to better adapt to actual needs. This method can more accurately reflect the actual situation, and at the same time reduce errors, providing a more efficient and safer flight area for UAV flight.

Description

Technical Field

[0001] The present invention relates to the fields of UAV technology and power system inspection, and particularly to a method for generating a UAV flight buffer zone based on homogeneous rules. Background Art

[0002] In the modern power industry, transmission lines cover a wide range and the environmental conditions are complex. Traditional manual inspection methods require inspectors to walk or climb on high altitudes or in dangerous terrains for inspections. This not only has a high labor intensity but also poses extremely high safety risks, especially in mountainous areas with steep terrains, high towers spanning rivers, and remote areas far from populated areas. This inspection method is inefficient and cannot meet the power industry's requirements for efficient and stable inspections.

[0003] To solve the above problems, UAV inspection technology has emerged. By carrying high-definition cameras, infrared imaging devices, and sensors, UAVs can quickly obtain high-resolution images and real-time data of transmission lines and their surrounding environments. Compared with manual inspections, UAVs have the advantages of large coverage, fast inspection speed, and low cost, significantly improving inspection efficiency and safety. When performing inspection tasks, UAVs must fly in a changing and complex environment, such as mountains, dense forests, buildings, and adverse weather conditions. These factors pose potential threats to the flight safety of UAVs. To ensure that UAVs can complete inspection tasks safely and efficiently, it is necessary to generate a flight buffer zone. A flight buffer zone is an optimized safe flight area that can help UAVs avoid obstacles and dangerous areas, maintain a stable flight path, and thus reduce the risk of collisions and data collection errors.

[0004] Traditional buffer zone generation algorithms usually rely on simple fixed-distance models and do not consider the complex situations in the environment. This method can be used in relatively simple environments, but in complex environments, it may cause the following problems:

[0005] 1. The buffer zone is too large: In areas with few obstacles or flat terrains, an overly large buffer zone wastes flight space and reduces the inspection efficiency of UAVs.

[0006] 2. The buffer zone is too small: In areas with complex terrains or dense obstacles, a fixed buffer zone cannot provide sufficient safety distance, increasing the risk of flight collisions.

[0007] 3. Poor environmental adaptability: It cannot be dynamically adjusted according to geographical features, resulting in a lack of adaptability of UAVs.

[0008] To solve the above problems and improve the safety and efficiency of UAV inspection, there is an urgent need to develop an intelligent algorithm for dynamically adjusting the buffer zone. The method for generating a UAV buffer zone based on homogeneous rules proposed by the present invention dynamically adjusts the shape and size of the buffer zone by analyzing the surrounding environment data, enabling the UAV to complete the inspection task more safely and efficiently. Summary of the Invention

[0009] The object of the present invention is a method for generating a UAV buffer zone based on homogeneous rules, which dynamically adjusts the shape and size of the buffer zone by analyzing the surrounding environment data, and improves the problem that the existing technology cannot be applied to complex environments.

[0010] A method for generating a UAV flight buffer zone based on homogeneous rules provided by the present invention includes the following steps:

[0011] Obtain the basic parameters of the power line from the information collected by the UAV host platform and the point cloud to generate a line model;

[0012] Determine its state according to the vector product of two vectors of adjacent points in the line model, and preset different rules according to the point position and type to generate a basic boundary;

[0013] Calculate the buffer zone using homogeneous rules according to the preset different rules, including:

[0014] Calculate the participation index PI to reflect the attribute similarity between adjacent sampling points;

[0015] Mine general homogeneous rules;

[0016] According to the basic boundary and the participation index PI, use homogeneous rules to correct the boundary;

[0017] Use cubic spline interpolation to connect the corrected points to form a continuous and smooth buffer zone boundary to obtain the buffer zone coordinates.

[0018] A method for generating a UAV flight buffer zone based on homogeneous rules provided by the present invention has the technical effects of strong environmental adaptability and easy implementation.

[0019] Optionally, the information collected from the UAV host platform and the point cloud includes the longitude, latitude and height of the tower, and the information is converted into projection coordinates to obtain the basic parameters of the power line and generate a line model.

[0020] Optionally, the step of determining its state according to the vector product of two vectors of adjacent points in the above line model includes:

[0021] Connect the projection coordinates of three adjacent towers to form two initialization line segments;

[0022] Determine the vector representations of two initialization line segments and calculate their vector products;

[0023] Judge the point attributes according to the vector product. If it is an inflection point, it is necessary to further judge its concavity and convexity.

[0024] Optionally, the step of presetting different rules according to the point position and type and generating a basic boundary includes:

[0025] When the point is neither an inflection point nor an endpoint, calculate the Euclidean distance between the point and a specific point located on the discovery. If the Euclidean distance is less than a given threshold, the point is a candidate that meets the homogeneity rule.

[0026] Optionally, the step of presetting different rules according to the point position and type and generating a basic boundary includes:

[0027] When the point is an inflection point, first draw two perpendicular lines along the adjacent two line segments, then draw parallel lines pointing to the adjacent two line segments, which intersect with the buffer boundary points. Then, construct an arc on the convex side with the inflection point as the center and the distance from the inflection point to the parallel line as the radius. The arc intersects with the two perpendicular lines to form a closed area, which is the initialization area for meeting the homogeneity rule.

[0028] Optionally, the step of presetting different rules according to the point position and type and generating a basic boundary includes:

[0029] When the point is an endpoint, draw a perpendicular line segment through the endpoint and generate parallel lines on the existing buffer boundary points, which correspond to the points closest to the endpoint. Then draw an arc with the intersection point of the perpendicular line segments as the center and the distance from the intersection point to the parallel line as the radius. The arc intersects with the two parallel lines respectively, and the perpendicular line segment and the arc form a closed area, which is the initialization area for meeting the homogeneity rule.

[0030] Optionally, the step of calculating the participation index PI to reflect the attribute similarity between adjacent sampling points includes:

[0031] Define the participation index formula;

[0032] Calculate the participation index of all points according to the participation index formula;

[0033] When The value is greater than or equal to 0.5, indicating that these instances meet the homogeneity rule;

[0034] When The value is less than 0.5, indicating that these instances do not meet the homogeneity rule.

[0035] Optionally, the step of mining the general homogeneity rule includes:

[0036] Determine whether the candidate homogeneous rule is a universal homogeneous pattern based on the participation index of all points;

[0037] If the PI is not equal to 1, it is considered that the candidate homogeneous rule is not a universal homogeneous pattern;

[0038] Continue the homogeneous pattern mining process until there are no qualified homogeneous patterns.

[0039] Optionally, the step of correcting the boundary using the homogeneous rule according to the basic boundary and the participation index PI includes:

[0040] Define the calculation formula for the buffer distance;

[0041] Calculate the distance between the sampling points of the target line segment and their corresponding buffer boundary points according to the formula;

[0042] Judge whether the universal homogeneous rule is satisfied according to the participation index PI to determine the width and shape of the buffer zone.

[0043] Optionally, the step of using the cubic spline interpolation method to connect the corrected points to form a continuous and smooth buffer boundary to obtain the buffer coordinates includes:

[0044] Use a cubic polynomial between every two adjacent points in a set of data points;

[0045] Determine the coefficients of the cubic polynomial through a linear equation;

[0046] Finally, convert the obtained coordinate information into "EPSG:4326" geographic coordinates and store them in the database. Description of the Drawings

[0047] Figure 1 is a flowchart of a method for generating a UAV flight buffer zone based on homogeneous rules provided by the present invention. Detailed Embodiments

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein shall have the ordinary meanings understood by those of ordinary skill in the art in the field to which the present invention belongs. The words such as "including" used herein mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects.

[0049] An embodiment of the present invention provides a method for generating a UAV flight buffer based on homogeneous rules, including the following steps:

[0050] S1. Obtain the basic parameters of the power line from the information of the UAV host platform and point cloud collection to generate a line model;

[0051] S2. Determine its state according to the vector product of two vectors of adjacent points in the line model, and preset different rules according to the point position and type to generate a basic boundary;

[0052] S3. Calculate the buffer using homogeneous rules according to the preset different rules, including:

[0053] S3.1. Calculate the participation index PI to reflect the attribute similarity between adjacent sampling points;

[0054] S3.2. Mine general homogeneous rules;

[0055] S3.3. According to the basic boundary and the participation index PI, use homogeneous rules to correct the boundary;

[0056] S4. Use cubic spline interpolation to connect the corrected points to form a continuous and smooth buffer boundary to obtain buffer coordinates.

[0057] The flowchart of a method for generating a UAV flight buffer based on homogeneous rules provided by the present invention is as Figure 1 shown.

[0058] When performing step S1, the tower information obtained according to the UAV system is initialized as a line segment, and the tower information includes the longitude, latitude and height of the tower. The target line tower object is virtually recorded as a series of points, and each tower is a point, so a series of points are obtained on the target line and represented as:

[0059] Initialize the line segment: First, obtain the basic parameters of the power line from the information of the UAV main station platform and point cloud collection, including tower attributes, and convert the position data into "EPSG:3857" projection coordinates to generate a line model.

[0060] In one embodiment, the target line tower object is virtually recorded as a series of points, and each tower corresponds to a point, so a series of points are obtained on the target line and represented as .

[0061] When performing step S2, determine the linearity of three adjacent points: Assume that the coordinates of three adjacent points are and , and their vectors are represented as , , specifically represented as:

[0062]

[0063] ;

[0064] The vector product of two vectors is used to determine their linearity, and the calculation formula is:

[0065] ;

[0066] where , , , ; is the unit vector. If equals zero, then and are collinear, that is, the adjacent three points are almost on a straight line; if is greater than or less than zero, the middle point of the three points is usually an inflection point. In addition, the concavity and convexity of the inflection point should be judged. If > 0, it is a convex point; if < 0, it is a concave point;

[0067] Perform different processing according to the point attributes to obtain the basic boundary information:

[0068] When the point is neither an inflection point nor an endpoint, assume is the normal line on the straight line , the instance point and the point located on this normal line. The Euclidean distance between them is

[0069] ;

[0070] where, is the coordinate of the point , is the coordinate of the point , if is less than or equal to the given threshold , then there is an adjacent relationship between A2 and Q2. Therefore, these points are considered candidates that meet the homogeneity rule.

[0071] When the point is at the inflection point, first draw two vertical lines along the adjacent two line segments, then draw parallel lines pointing to the adjacent two line segments, which intersect with the buffer boundary points. Then, construct an arc on the convex side with the inflection point as the center and the distance from the inflection point to the parallel line as the radius. The arc intersects with the two vertical lines to form a closed area, which is the initialization area that meets the homogeneity rule.

[0072] When the point is located at the endpoint, draw a vertical line segment through the endpoint and generate parallel lines at the existing buffer boundary points, which correspond to the points closest to the endpoint. Then draw an arc with the intersection point of the vertical line segment as the center and the distance from the intersection point to the parallel line as the radius. The arc intersects the two parallel lines respectively. The vertical line segment and the arc form a closed area, which is the initialization area for satisfying the homogeneity rule.

[0073] When performing step S3.1, first determine the rough candidate area using the Euclidean distance. The Euclidean distance is usually the most commonly used calculation method and is defined as:

[0074] ;

[0075] where are the coordinates of the sampling point A2, are the coordinates of the corresponding buffer boundary point.

[0076] The determination and adjustment of the buffer distance depend on the homogeneity rule. The use of the homogeneity rule requires the use of the participation index , and the participation rate is an index used to measure the similarity between different instances or samples. In the homogeneity rule, the participation rate is mainly used to determine whether the attributes of multiple sampling points are similar within a specific area, so as to decide whether these instances can be regarded as homogeneous.

[0077] The participation rate is usually determined by calculating statistical quantities such as the mean and variance of a specific attribute of a group of instances. The basic idea is to evaluate the consistency of these instances in this attribute.

[0078] In one implementation method, assume there is a group of instances whose values on a certain attribute are respectively , and the participation rate can be defined as:

[0079] ;

[0080] where, is the mean of attribute X;

[0081] where, is the standard deviation of attribute X;

[0082] where, is the total number of instances.

[0083] When is close to 1, it indicates that the attribute similarity between instances is relatively high, indicating that they are suitable under the homogeneity rule.

[0084] When When the value is close to 0 or negative, it indicates that there are significant differences in attributes between instances, and these instances may not meet the homogeneity condition.

[0085] When performing step S3.2, when all points use the participation index ( ), candidates for instances that meet the homogeneity rule are determined. If PI is not equal to 1, it is considered that this candidate homogeneity rule is not a universal homogeneity pattern. This process of mining homogeneity patterns will continue until there are no qualified homogeneity patterns.

[0086] When performing step S3.3, the calculation formula for the buffer distance is:

[0087] ;

[0088] where D is a function for calculating distance, is the homogeneity rule that determines the adjustment of the buffer distance, reflecting the similarity of the attributes of sampling points in the surrounding environment.

[0089] The function D is used to calculate the distance between the sampling points of the target line segment and its corresponding buffer boundary points The main purpose is to determine the actual distance from each sampling point to its corresponding boundary point during the buffer generation process, so as to provide basic data for dynamically adjusting the buffer distance;

[0090] This rule determines the width and shape of the buffer by analyzing the attributes of neighboring instances to judge whether they meet certain homogeneity conditions. If the adjacent sampling points and have high attribute similarity, it indicates that these areas have similar environmental characteristics, and the buffer can maintain a larger width; otherwise, the buffer may need to be narrowed.

[0091] When performing step S4, after all buffer boundary points are determined, the cubic spline interpolation method is used to connect these points to form a continuous and smooth buffer boundary. The boundary line will finally form a closed area to provide safe path protection for the UAV flight.

[0092] In one embodiment, assume there is a set of data points . A cubic polynomial is used between every two adjacent points:

[0093] ;

[0094] where, , are the corresponding coefficients. Their values are determined by a system of linear equations, and the following is the calculation method:

[0095] First, the curve must pass through each interpolation point, and at the points and find :

[0096] ;

[0097] These two formulas determine and are partially related to .

[0098] Continuity of the first derivative. The slope (derivative) of the curve at each internal node is continuous:

[0099] ;

[0100] After expanding the formula, it involves .

[0101] Continuity of the second derivative. The curvature (second derivative) of the curve at each internal node is continuous:

[0102] ;

[0103] After expanding the formula, it involves .

[0104] Boundary conditions

[0105] Boundary conditions are needed to determine the start and end points:

[0106] Natural boundary conditions (most commonly used): Assume that the second derivatives at both ends are 0:

[0107] ;

[0108] According to all the above coefficient formulas, transform them into a system of linear equations. By solving this system of equations, all coefficients can be calculated.

[0109] Finally, convert the obtained coordinate information into "EPSG:4326" geographic coordinates and store them in the database.

[0110] Although the embodiments of the present invention have been described in detail above, it is obvious to those skilled in the art that various modifications and changes can be made to these embodiments. However, it should be understood that such modifications and changes are all within the scope and spirit of the present invention described in the claims. Moreover, the present invention described herein can have other embodiments and can be implemented or realized in various ways.

Claims

1. A method for generating a flight buffer zone for a UAV based on homogeneous rules, characterized in that: The following steps are involved: The basic parameters of the power line are obtained from the information collected by the UAV host platform and point cloud to generate a line model; The state of adjacent points in the line model is determined according to the vector product of two vectors, and different rules are preset according to the point position and type to generate a basic boundary; the step of preset different rules according to the point position and type to generate a basic boundary includes: When a point is neither an inflection point nor an endpoint, the Euclidean distance between the point and a specific point on the discovery is calculated. If the Euclidean distance is less than a given threshold, the point is considered a candidate that satisfies the homogeneity rule. When the point is an inflection point, first draw two perpendicular lines along the two adjacent line segments, then draw parallel lines pointing to the two adjacent line segments and intersect with the buffer boundary point, then construct an arc on the convex side with the inflection point as the center and the distance from the inflection point to the parallel line as the radius. The arc intersects with the two perpendicular lines to form a closed area, which is the initialization area used to satisfy the homogeneity rule; When the point is an endpoint, draw a perpendicular line segment through the endpoint and generate parallel lines on the existing buffer boundary points. These points correspond to the points closest to the endpoint. Then draw an arc with the intersection of the perpendicular line segments as the center and the distance from the intersection to the parallel line as the radius. The arc intersects with the two parallel lines respectively. The perpendicular line segment and the arc form a closed area, which is the initialization area used to satisfy the homogeneity rule. Calculating the buffer zone using homogeneous rules according to the preset different rules, including: calculating the participation index PI to reflect the attribute similarity between adjacent sampling points, mining the universal homogeneous rules, and modifying the boundary using the homogeneous rules according to the basic boundary and the participation index PI; The step of calculating the participation index PI to reflect the attribute similarity between adjacent sampling points includes: Define a participation index formula; calculate the participation index of all points according to the participation index formula; when The value of is greater than or equal to 0.5 indicates that these embodiments meet the homogeneity rule; when If the value of is less than 0.5, these instances do not satisfy the homogeneity rule; The step of mining universal homogeneous rules includes: judging whether the candidate homogeneous rule is a universal homogeneous pattern according to the participation index of all points; if the PI is not equal to 1, it is considered that the candidate homogeneous rule is not a universal homogeneous pattern; and continuing the homogeneous pattern mining process until there is no qualified homogeneous pattern; The step of modifying the boundary using the homogeneity rule according to the basic boundary and the participation index PI includes: defining a calculation formula for the buffer distance; calculating the distance between the target line segment sampling point and its corresponding buffer boundary point according to the formula; judging whether the universal homogeneity rule is satisfied according to the participation index PI to determine the width and shape of the buffer zone; The corrected points are connected using cubic spline interpolation to form a continuous and smooth buffer boundary to obtain the buffer coordinates.

2. The method for generating a flight buffer zone for a UAV based on homogeneous rules according to claim 1 is characterized in that: The information collected from the drone host platform and the point cloud includes the longitude, latitude and height of the tower. The information is converted into projection coordinates to obtain the basic parameters of the power line and generate a line model.

3. The method for generating a flight buffer zone for a UAV based on homogeneous rules according to claim 1 is characterized in that: The step of determining the state of adjacent points in the line model according to the vector product of two vectors thereof comprises: Connect the projection coordinates of three adjacent towers to form two initialization line segments; Determine the vector representation of the two initialization line segments and calculate their vector product; The point attributes are determined based on the vector product. If it is an inflection point, its concavity and convexity need to be further determined.

4. The method for generating a flight buffer zone for a UAV based on homogeneous rules according to claim 1, characterized in that: The buffer coordinates obtained by connecting the corrected points using the cubic spline interpolation method to form a continuous and smooth buffer boundary include: Use a cubic polynomial for every two adjacent points in a set of data points; Determining coefficients of the cubic polynomial by linear equations; The final coordinate information is converted into "EPSG:4326" geographic coordinates and stored in the database.

Images